Lidar-Based Landing Hazard Detection for Dragonfly

NASA's Dragonfly mission will use a dual-quadcopter lander to explore the surface of Saturn's moon Titan. After Titan arrival, the entry, descent and landing sequence includes a transition from parachute descent to powered flight, autonomous selection of a safe landing site, and autonomous landing. Additionally, the lander will relocate to multiple sites of scientific interest on the Titan surface via powered flight. Due to the long light-time delay to Earth, limited flight time, and limited pre-existing map data for Titan, Dragonfly must utilize autonomous precision landing and hazard avoidance technology to search for safe landing sites. The Dragonfly team has developed a hazard detection approach based on imaging lidar data accumulated during powered flight above the Titan surface. This paper describes the development of the Lidar Terrain Sensing algorithms, which have been designed to provide robust hazard detection and safe site identification capability. Algorithm components include compiling a detailed topographic map of the surface, estimating the surface slope and roughness, assessing the relative safety of all map points, and suggesting the highest-scoring locations as candidate landing sites. The safety assessment is designed to include probabilistic estimates of sensor and algorithm noise effects, the lander mechanical tolerance to surface features, and navigation uncertainty from site selection to landing. The interaction of the Lidar Terrain Sensing algorithms with other onboard algorithms and mission operations will be discussed, along with algorithm performance in simulated flight conditions.